Linear fits to 2010–2018 X-ray rates predict too little spin-down in 4U 1626−67 during 2023–2025

Can an early relation between X-ray brightness and spin change predict the later spin of 4U 1626−67? We fit a straight line to public Fermi GBM frequencies and separate Swift/BAT and MAXI/GSC rates in 2010–2018. We integrate each brightness history between frequency measurements. We then keep the fitted relation fixed. In the check from 22 April 2023 to 14 October 2025, the measured mean frequency rate is −3.776 × 10^−13 Hz/s. BAT predicts +0.282 and MAXI predicts −0.337 in the same units. The accumulated errors are −31.79 and −26.94 microhertz. The earlier 2019–2022 check has smaller errors. All later interval brightness values are below the training range. The result is thus a failed extrapolation of these simple fits. It does not isolate a new physical change at the known 2023 reversal. The MAXI prediction changes less than the BAT prediction across the tested averaging methods. Error bands depend on assumptions about timing noise and monitor calibration. Larger assumed noise permits agreement. A short MAXI fit after a camera change also permits agreement. The package preserves the source data, hashes, interval tables, code, figures, and numerical checks.

4U 1626-67 · accreting pulsars · torque reversal · Fermi GBM · Swift BAT · MAXI GSC · prediction · timing noise · public data